Sticky Words? Towards a Theory of Rhetorical Path Dependency Dr

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Apollo

Sticky Words? Towards a Theory of Rhetorical Path Dependency

Dr Dennis C. Grube University of Cambridge Email: [email protected]

Abstract: Speech matters. Political actors are defined by what they say as much as by what they do. But with each rhetorical choice, they also narrow the range of rhetorical options open to them for the future. This paper examines the idea of path dependency, a well-established concept in the field of policy studies, and applies it to the study of political rhetoric. It argues that words are sticky, leaving political leaders caught between the desire to utilise fresh and engaging rhetoric to explain new policy choices and the reality that they can’t shake off the wording of their previous promises. In advancing a theory of rhetorical path dependency, the paper builds on the insights of both discursive institutionalism and rhetorical political analysis to suggest that whilst ideas are indeed vital to the shaping of institutions, the arguments that give those ideas shape can themselves be constrained by earlier choices.

Keywords: rhetoric; path dependency; discursive institutionalism; Abbott government;

Acknowledgements The author would like to acknowledge the support and encouragement of colleagues in the Institute for the Study of Social Change at the University of Tasmania, and the Centre for Governance and Public Policy at Griffith University.

Sticky Words? Towards a Theory of Rhetorical Path Dependency

Political leaders like to portray themselves as straight talkers. They understand the electoral benefits that can attach to those who are seen by the electorate as being honest and straightforward in their choice of words. Amidst a political discourse framed by the careful political marketing of politicians and party ‘brands’, individual political actors are fond of stressing that they are different; that they will cut through with some straight talk where others have simply dissembled. But the rhetorical choices of political actors frequently have longer-term consequences. Today’s killer line can evolve into tomorrow’s rhetorical dead weight when political actors are confronted by changing circumstances that don’t yield to the simplicity of their original rhetorical formulation.

This article argues that words are sticky. It draws on the concept of path dependency (Pierson 2000; 2004) as a way of analysing the longer-term effects of the rhetorical choices that political actors make. Rhetorical choices are examined not simply as a tool of political persuasion, but as promises that can assume institutional characteristics capable of limiting subsequent rhetorical options. The article argues that political leaders are caught between wanting to embark on new policy directions and the need to remain trustworthy in the eyes of the electorate by staying true to their earlier rhetorical formulations. Thus, the rhetorical choices made at critical junctures can define the range of future options available. Actors effectively become trapped in gilded rhetorical cages of their own making, wherein the very success of their earlier rhetoric paradoxically prevents them from easily adopting new rhetorical formulations even when circumstances may require it. The article seeks to investigate the theoretical underpinnings of this phenomenon, and identify the factors that can work to aggravate or mitigate rhetorical path dependent effects.

In the sections that follow, the article begins by setting out existing theories of path dependency and how they can be usefully applied to the study of rhetoric. It then examines some of the more recent theoretical advancements in the study of ideas, discourse and rhetoric. Specifically it addresses discursive institutionalism (Schmidt 2010) and rhetorical political analysis (Finlayson 2007) to suggest that whilst ideas are indeed vital to the framing of institutions, the arguments that give those ideas shape can themselves be constrained by earlier choices. The paper sets out a heuristic of the factors likely to impact on the degree of path dependency that may attach to any particular rhetorical formulation, and sets out a short exploratory case study to demonstrate the heuristic’s applicability. The final section suggests areas for further research, and examines the potential and the limitations of viewing political rhetoric through this kind of path dependent frame.

Path Dependency and Discursive Institutionalism Path dependency is a concept that emerged from theories of historical institutionalism, arguing for the need to take what one might call the ‘long view’ when assessing policy processes and institutional shapes and how they might have come about. In essence, path

3 dependency argues that past action has significant consequences for the range of future options that any individual, organisation or institution has available to them (see Mahoney and Thelen 2010; Pierson 2004). ‘Path dependency speaks to the common observation that the legacy of the past conditions our future, at the policy level and at the level of institutions’ (Gains, John, and Stoker 2005: 27). Kay (2005), drawing on North (1990), argues that a ‘…process is path dependent if initial moves in one direction elicit further moves in that same direction; in other words the order in which things happen affects how they happen; the trajectory of change up to a certain point constrains the trajectory after that point’ (2005: 553).

In conceptualising path dependence, Mahoney (2000) emphasises that path dependency is about more than simply saying that ‘history matters.’ He argues that ‘path dependence characterizes specifically those historical sequences in which contingent events set into motion institutional patterns or even chains that have deterministic properties’ (2000: 507). In other words, the point at which an initial choice or decision is made – often characterised as a ‘critical juncture’ (see Hogan 2006) – creates boundaries that limit future choices. Mahoney identifies two types of analysis practised by path dependency scholars. The first is to focus on ‘self-reinforcing sequences’, in which ‘increasing returns’ entrench patterns of behaviour because of the costs associated with going back to the start to begin anew (Mahoney 2000: 508). The second is to examine ‘reactive sequences’ in which events are studied as a sequential chain, with each link being a reaction to the one preceding it, with each forming a step in the path that leads to whatever outcome is being studied (Mahoney 2000: 509). In both forms of analysis a ‘path’ is a pattern of continuing behaviour, with the difference being in the causal forces shaping the pattern.

Path dependency does not argue in any of its forms that change never occurs, merely that it is difficult. And its key insight in this regard is that the costs of exit rise as the benefits of continuity grow, making change not just difficult but costly in either a fiscal or a political sense. As a result, change only occurs in a very wrenching fashion when a critical juncture arises. Critiques of path dependency – as with critiques of historical institutionalism more broadly – have focussed on allegations that it is overly- deterministic and doesn’t allow sufficiently for both the influence of individual agency and the impact of incremental change in between the critical junctures that jolt a system off its prior path (see Peters, Pierre, and King 2005; Schmidt 2010). As Kay argues, the view that critical junctures can only take the form of a single exogenous shock – akin to a big bang – is too binary, and we must allow for the possibility that cumulative endogenous change can have a sufficient impact in potentially changing the direction of a policy path (Kay 2003: 416).

Path dependency’s insight that the cost of changing paths rises over time is particularly useful when examining its effects on political rhetoric. A piece of political rhetoric uttered only once in passing is less likely to attract strongly path dependent effects than something which is repeated and consistently adhered to over time. Pierson argues that this issue of ‘time’ is central to understanding how path dependency works:

Increasing returns dynamics capture two key elements central to most analysts’ intuitive sense of path dependence. First, they pinpoint how the costs of switching from one alternative to another will, in certain social contexts, increase markedly over time. Second, and related, they draw attention to issues of timing and sequence, distinguishing formative moments or conjunctures from the periods that reinforce divergent paths. In an increasing returns process, it is not only a question of what happens but also of when it happens. Issues of temporality are at the heart of the analysis. (2000: 251)

Actors who are experiencing the increasing returns of staying on a path are of course less inclined to accept the costs of ultimately deciding to stray from it (Gains et al 2005: 27).

I argue here that these same path dependent effects that are discernible in the making and changing of public policy extend equally to rhetorical choices. If an actor is experiencing positive benefit from a particular rhetorical formulation, they will be disinclined to change tack, even if a change in circumstances suggests that they should. But more importantly, in the process of experiencing those positive benefits they are highly likely to have entrenched the grip of the rhetorical formulation within the public discourse through re-iteration, emphasis, and repetition. And the more embedded a particular piece of rhetoric has become in the discourse, the more path dependent its effects in terms of the political or electoral costs that will follow from any repudiation of it.

The ways in which rhetorical path dependency constrains actors relies on a dynamic conceptualisation of path dependency that recognises the key part that individual agents play in discursively erecting the very boundaries that end up restraining them. In essence, decisions endogenous to political actors – in response to assessments of potential exogenous feedback – both begin the process of rhetorical path dependency, and end it once it becomes clear that a particular rhetorical path is simply no longer tenable. The process begins with the exercise of individual choice. A political actor gives voice to a particular rhetorical formulation to describe a policy problem or their own approach to a political issue. That rhetorical choice is then pursued and reinforced through talking points and press interviews. At each point, through repetition and honing, the initial rhetoric is further entrenched. In effect, rhetorical actors paint themselves into a corner with repeated brushstrokes that emphasise their position. When faced with the need to break-out of that corner, they find instead that they have trapped themselves by linking their public reputation to staying exactly where they are. So what has trapped them is the institutional shape that they themselves have given to their words by the way they have framed and repeated them in the process of setting out their policy positions.

The capacity for discourse to shape and re-shape institutions has received significant attention over the past decade through the emergence of what has been titled ‘discursive institutionalism’ (Schmidt 2008; 2010; 2012; see also Hay 2010). Discursive institutionalism (DI) brings together under one umbrella those scholars who argue for the centrality of ideas in building and shaping institutions. The catalyst for the emergence of discursive institutionalism was a sense of frustration with the alleged shortcomings of the

5 three older ‘new institutionalisms’ – rational choice (Hardin 1982; Shepsle 2008), historical (see Bell 2011), and sociological (Mackay, Monro, and Waylen 2009; March and Olsen 1989) – at being able to adequately explain institutional change.

Discursive institutionalism suggests that there are two spheres which work together to shape and change institutions; the back of mind ideas about how things are and should be and the front of mind capacity to then communicate those ideas by framing them in a particular fashion (Schmidt 2010: 4). As Schmidt argues, DI is used to demonstrate how ‘discursive interactions enable actors to overcome constraints which explanations in terms of interests, path dependence, and/or culture present as overwhelming impediments to action’ (Schmidt 2010: 4). DI privileges agency to the extent that it allows for actors to change institutions by re-shaping the ideas that underpin them and then ‘bring the ideas developed in the context of the coordinative discourse to the public for deliberation and legitimation’ (Schmidt 2010: 3). In this formulation, discourse retains a free hand in shaping institutions, limited only by the capacity of actors to make a compelling case when advocating for change.

Rhetorical path dependency acts as a corrective to the conceptual starting point that communicative behaviour is itself unrestrained in its potential freedom of action. Rhetorical path dependency suggests that there are significant restrictions on how dynamic language can be, based on the capacity of the speaker to project persuasive and authentic arguments without contravening their own earlier rhetorical formulations. Rhetorical path dependency is not deterministic; it does not suggest that words are unchangeable, but it does argue that they are sticky. Whilst discourse might theoretically be fluid amongst multiple actors, individual agents can in fact erect new institutional barriers for themselves even as they seek to re-shape older ones. In order to examine how a theory of rhetorical path dependency might interact with previous work on rhetoric, the next section examines the existing literature on political rhetoric and how it might accommodate new theoretical perspectives.

From Rhetoric to Rhetorical Path Dependency There is a long tradition of US scholarship on the role of rhetoric as a tool of persuasion in a system where there is a true separation of powers. Seminal works by Tulis (1987) and Kernell (2006) have engaged with the ways that the presidency in particular can use its rhetorical power to reach out directly to the public over the heads of a fractious congress or a combative supreme court. This kind of analysis spurred a deep and continuing scholarship on the power and role of presidential rhetoric that shows little sign of slowing down (Friedman and Friedman 2012; Medhurst 2006; Stuckey 2010). Until the last decade or so, no commensurate body of scholarship had emerged to analyse the role and impact of political rhetoric within parliamentary systems of government operating under the Westminster model. Important recent works have begun to address this lacuna. There have been studies of the relationship between rhetoric and leadership, including theoretical work and comparative work in the United Kingdom and Australia, and on the role of rhetoric in shaping political debates and perceptions of leaders (Atkins, Finlayson, Martin and Turnbull 2014; Bennister 2013; Grube 2013; Finlayson and Martin

2008; Kane and Patapan 2010; Martin 2015; Rolfe 2008; Toye 2011; Uhr and Walter, 2014a; 2015; Young 2007).

Collectively, this emerging body of scholarship has highlighted the complex range of roles that rhetoric plays in the contemporary body politic. It demonstrates that political actors cannot avoid the difficulties of making rhetorical choices because it is an inherent part of their institutionalised role in a democracy – it is at the core of their vocation (see Uhr and Walter 2014b: 243). As Finlayson and Martin argue, organised speech moments ‘…are institutionalised events with a tradition, ritualistic and generic character; spaces within which roles are arranged and ritually enacted so as to confirm their existence and test the competence of individuals to fulfil them’ (2008: 447). We expect our leaders to speak publicly and persuade us when they do so – it is part of the electoral test of whether they are sufficiently competent to be elected or re-elected.

The work of Finlayson, together with co-authors such as Atkins and Martin, has argued for a new ‘Rhetorical Political Analysis’ (Atkins and Finlayson 2013; Finlayson 2007; Finlayson and Martin 2008). At the core of Rhetorical Political Analysis (RPA) is the argument that there is more to rhetorical communication than the words that are spoken, because the rhetoric is at one and the same time constitutive and formative of the ideas that are expressed through those words. And, importantly, ideas can only be accessed by studying the arguments made for or against them by political actors employing political rhetoric. So in studying rhetoric, we are in reality studying the creation, shaping and re- shaping of political ideas, through the arguments that are being made for or against those ideas. The study of rhetoric becomes, instead of the close study of linguistics, a study of politics itself. To quote Finlayson:

If we begin with a clear and distinct concept of politics as the ‘arena’ within which we see expressed the irreducible and contested plurality of public life, the ineradicable contestation of differing world-views, then it is clear that what is distinct in politics is not the presence of beliefs but the presence of beliefs in contradiction with each other, not decisions about courses of action but of dispute over decisions and courses of action. It then follows that ideational and interpretive analyses have tended to examine the wrong object, which ought to be not ideas but arguments: their formation, effects and fate in the activity known as persuading. The study of such argument and persuasion necessitates the development of Rhetorical Political Analysis. (Finlayson, 2007: 552)

Finlayson and his co-authors are surely right in centring political arguments as a defining feature of what politics actually is. Without arguments – without the power to persuade others through rhetorical contestation – we are left with a kind of discursive stasis in which individual actors have to wait for the impact of exogenous factors beyond their control to open new political opportunities for them. But important limitations remain on what sort of arguments can legitimately be made at different points by different actors. A theory of rhetorical path dependency builds on the insights of RPA by drawing attention to the ways in which rhetoric contributes not just to the shaping of political ideas through

7 arguments, but also to building discursive structures that can have important effects in shaping subsequent political action. In other words, rhetorical path dependency suggests that the power of rhetoric to shape and transmit political ideas through arguments is not unrestrained. Past arguments used by political actors in particular ways may limit the choice of future arguments available to them. Political actors do not start with a clean sheet of paper each time that they formulate a piece of political rhetoric – their rhetorical choices are encumbered by the ways they have shaped their rhetoric previously. What’s more, the costs associated with deviating from an established rhetorical path impact directly on the level of trust the public is prepared to invest in a leader.

Empirically, there is a link between the consistency of a leader’s political rhetoric and public perceptions of their authenticity as a leader (Grube 2013: 18-19). This in turn contributes to whether or not a particular actor is seen as authoritative and worthy of the public’s trust. In order to understand why this might be so, scholars of political rhetoric still begin with the insights of Aristotle (Aristotle trans, 1984). In Aristotelian terms, what politicians say simultaneously shapes and reflects their ‘ethos’, which has a large impact on how persuasive their rhetoric is able to be. For Aristotle, the three modes of rhetorical persuasion were logos, pathos and ethos (see Finlayson 2012). Logos speaks to the inherent power of the argument through the proofs and evidence put forward. Pathos refers to the emotional connection that a speaker is able to elicit from the audience.

Ethos is in some ways the most complicated of the three components because it is unable to be manufactured as part of the speech act itself. Ethos relies at least in part on a pattern of behaviour that has led the audience to view the speaker as authoritative and worth listening to. At its core, ethos is about trust. Voters are more likely to trust and believe the rhetoric of a leader who is seen as consistent in what she or he says, rather than being perceived as blowing with the wind of political fortune to select the rhetoric that they think is the most popular at any given point in time. This then begins to shape and frame the public face of the speaker in a kind of recursive action. The more they repeat and emphasise their rhetorical points, the more those words are entrenched in the public mind as being attached to that particular actor, which in turn increases their reputation for consistency.

The difficulty for politicians is that circumstances inevitably change. Modern governance is complex and what seemed like very sound rhetorical framing of policy positions can be quickly overwhelmed by what former British Prime Minister Harold Macmillan characterised as ‘events, dear boy, events.’ Faced with changing events, leaders will inevitably have to make new rhetorical choices about how to frame their policy positions. But, ironically, the more successful they were in entrenching their previous rhetorical choices into the public mind, the higher the political costs that will be attached to them suddenly moving to a different formulation. What was once a highly successful rhetorical strategy – such as categorically ruling out tax changes – can quickly become a dead weight that a leader is unable to shake off. The costs of change become prohibitive, even as the necessity for change becomes irresistible.

Rhetorical Path Dependency: A Heuristic and An Exploratory Case Study

In outlining the restraining effects of rhetorical path dependency above, I have argued that rhetoric can be sticky, but that change is possible. There remains room for actors to exercise agency in re-shaping their earlier rhetoric to try and re-frame a particular issue in a way that at least provides the illusion of consistency. Clearly not every word a political leaders utters will attract the same level of stickiness in immediately creating path dependent effects. To analyse the impact and degree of rhetorical path dependency, it is necessary to identify the variables that contribute to the stickiness of particular rhetorical formulations in particular situations. In the heuristic below, I suggest six variables that are likely to effect the degree of path dependency that attaches to particular instances of political rhetoric. These are summarised in tabular form in Figure 1. The list is suggestive rather than exhaustive, and like all heuristics its worth will be assessed over time as it is tested against more examples, and further potential variables are identified.

I then briefly apply this heuristic to a short exploratory case study to assess how each factor potentially contributes to the degree of rhetorical path dependency achieved in a particular instance. Following Eckstein (1975) and Flyvbjerg (2006), I utilise the exploratory case study approach as part of the process of theory building and testing to build a picture of the parameters within which rhetorical path dependency might operate. I select as my case the rhetoric used by the Abbott government around the idea of a ‘budget emergency’. As a case, it represents a situation where it was not the economic circumstances that changed, but rather the political context around those circumstances which forced the government to rhetorically re-frame the policy position it had adopted.

The Heuristic I argue that the six factors identified below contribute to how deeply entrenched a particular piece of rhetoric becomes in the public mind, and therefore how much it is likely to act as a restraint on political actors when they want to change their rhetoric in response to changing situations. Each of these six factors have attracted significant scholarly interest from different parts of the political science discipline, but have not been examined in terms of their ability to contribute to a form of rhetorical path dependency.

1. Political saliency (see Budge 2015): - It is a truism to say that voters do not weight the importance of all issues equally. And even if voters feel that an issue is important, saliency is ultimately a measure of whether it is important enough to change someone’s vote, and whether voters care enough to closely follow rhetorical fluctuations in how political actors are framing an issue. In terms of path dependent effects, the higher the level of saliency associated with an issue, the stickier the rhetoric a political actor uses is likely to be.

2. Proximity to the election: - elections are the most fiercely contested arena in which political actors compete for the public’s attention and affection. They are also the moment when the greatest proportion of the electorate are likely to be the most closely engaged in paying attention to the rhetoric of political leaders. The things that are said in the lead-up to an election are likely to gain the status of an ‘election promise’ (see Naurin 2014), raising the bar for the degree to which any rhetorical

retreat later will be seen as a breach of trust with the electorate. The closer to an election something is said, the stickier the words are likely to become.

3. Level of contingency and control: - Under doctrines of ministerial responsibility – and democratic accountability more broadly – voters hold their elected representatives responsible for what happens on their watch. Part of that process involves electors forming judgements over whether events really have been unforseen and have forced political leaders into having to change path, or whether those leaders had it in their power to control events and have failed to do so. As a result, leaders seek to frame their actions in certain ways to minimise blame but maintain their authority (Masters and ‘t Hart 2014: 761; see also Hood 2011). The less contingent on circumstances a promise is, the higher the path dependent effects are likely to be because political actors will be held to be in control and therefore responsible for how events have turned out.

4. Depth of the rhetorical commitment: - The depth of rhetorical commitment seeks to capture a number of related factors. These include the conviction with which the rhetorical commitment was first made (i.e. was it a throwaway line or the centrepiece of a policy speech), and the extent to which that commitment has then been hammered home by repetition or subsequent support. In other words, it captures the effects of time, with the longer a rhetorical message has been adhered to, the greater the path dependent effects likely to be attached to it. Equally, the more a rhetorical commitment is presented as a signature policy, and the more it is repeated, the higher the level of path dependency that it will engender.

5. Specificity of the rhetorical formulation: - Political leaders are often accused by commentators of using ‘weasel words’ to avoid being pinned down to a specific commitment. In effect, politicians are attempting to create wriggle room for themselves so that they can legitimately change path in response to changing circumstances if need be, without being hampered by the path dependent stickiness of their earlier rhetorical commitments. The more specific a rhetorical formulation is, the more likely it is to attract path dependent effects. For example, the nomination of a specific quantitative target for employment levels, or carbon emissions, or any aspect of public policy, is likely to be stickier than a general commitment to broad goals in these areas.

6. Media Coverage: - In the context of a hyper-vigilant media operating in a 24/7 news environment, the capacity for a piece of rhetoric to become sticky is heavily influenced by how it is framed and re-broadcast by the media. If a political or policy story is reported by the media in a way that focuses on a particular form of words that a political actor used or promise that was made, those are the words most likely to stick. Of course, political actors are well aware of this and tailor their rhetoric to fit with the needs of the modern ‘sound bite’.

These six factors together work to determine the cost of exit from the rhetorical path chosen by a political leader. If all six factors are present and ‘dialled up high’ as it were, the political cost associated with breaking from a rhetorical path becomes extremely prohibitive. Where only some factors are present, or are present at a lower level, the costs of rhetorical exit may in fact be quite easily politically managed. Individual agency therefore remains an important contributing factor in determining how much political damage results from re-setting rhetorical formulations to start again.

INSERT FIGURE 1 HERE

Case Study: Overcoming the Budget Emergency In the September 2013 Australian federal election the minority government led by Labor Prime Minister Julia Gillard was defeated by the incoming coalition government, led by Liberal Party leader Tony Abbott. Abbott won a comfortable majority in the lower house, but failed to secure control of the Senate, where an array of minor parties were successful in winning seats. Notably, the Palmer United Party – formed under the leadership of the mining magnate Clive Palmer – won three Senate seats, making them crucial to the balance of power.

In the lead up to the election, and then again in the lead up to the government’s first budget in 2014, Abbott and his Treasurer Joe Hockey adopted a rhetorical position which sought to persuade the Australian electorate that they faced a ‘budget emergency.’ Confronted with stubbornly high levels of government debt following the global financial crisis, (although at considerably lower levels of ratio to GDP than most other OECD countries), the government stressed that it had to act. It was a consistent refrain that ran through the rhetoric of Tony Abbott in particular, both as opposition leader and then as prime minister. In his budget reply speech as opposition leader in 2013, Abbott lambasted the Labor Government for its economic management and declared that Australia now faced a budget emergency, weighed down by ever mounting debt.

The government promised a surplus over the cycle but this isn’t a cycle – it’s a spiral, deeper and deeper into debt which is now surging towards $400 billion even on the government’s own figures…thanks to Labor’s poor management over five years, there is now a budget emergency. (Abbott 2013)

In response to what they it rhetorically identified as a ‘budget emergency’, the Abbott Government’s 2014 budget introduced a range of proposed cuts in government spending and changes to programmes. Declaring that the ‘age of entitlement’ was over, Treasurer Hockey asserted the need to act in the face of the crisis inherited from Labor: ‘On the back of five budget deficits in a row we have inherited a further $123 billion of deficits and debt rising to $667 billion. This challenge is not of our making, but we, the women and men behind me, accept responsibility to fix it. Doing nothing is not an option. The days of borrow and spend must come to an end’ (Hockey 2014). The budget crucially failed to win over key Senate cross-benchers, including senators from the Palmer United Party. Over subsequent months, the government’s support in the polls began to decline

11 steeply in the face of the proposed budget cuts, which emboldened the Senate to reject many of the proposed measures. The Government was forced to back down on heavily contested policies such as introducing a Medicare co-payment for each visit to the doctor, a new paid parental leave scheme, and cuts to the government funding of universities.

In the face of Senate blockages, and to combat negative public perceptions as measured in consistently poor opinion polls, the government sought to perform a rhetorical switch in March 2015. Instead of being in crisis, the budgetary situation was now declared by the prime minister to be under control: ‘There’s no cause for alarm under this government because we have got the budget situation from out of control to manageable’ (cited in Crowe and Uren 2015). Despite much of that previous budget not having been implemented, Abbott suggested that the need for harsh cuts was over, and that the 2015 budget would be ‘dull and boring’. He argued that the 2014 budget had done most of the necessary heavy lifting, and that as a result ‘this Budget certainly will be much less exciting than last year's Budget because the task this year is at least 50 per cent reduced from the task last year’ (Abbott, 2015).

Abbott was heavily criticised by media commentators for constructing a rhetorical U-turn that was not seen as credible, because of his earlier rhetoric. Having spent many months driving home a consistent message around a ‘budget emergency’, Abbott found that starting afresh in a new rhetorical direction was not easily achieved. His new rhetorical frame asserting that things were now under control was inevitably assessed through the lens of the fervour that had accompanied the original message. The earlier message remained sticky, and had operated to trap the government onto a rhetorical path that was simply no longer electorally viable, given the widespread opposition to the government’s 2014 budget. But the political cost of changing from that path proved to be prohibitive, leading to further political pain in opinion polls, and further damage to Abbott’s ethos – potentially damaging his ability to promulgate credible rhetorical changes in direction for the future.

In essence, the underlying ‘idea’ being promulgated by the government remained the same. This was the belief that balanced budgets matter, and that Labor governments could not be trusted to deliver them, whereas Coalition governments could. This idea was itself part of a wider context, with a long history of contestation between the parties as to who was the better economic manager (see Fenna 2007; Walter and Uhr 2013).

Applying the arguments of Finlayson referred to earlier, this kind of underlying idea can best be accessed by studying the arguments promulgated for and against it. Abbott wanted to continue to present the Coalition as the better economic manager, but was forced into changing his rhetorical arguments in support of that idea from saying that there was a budget emergency to saying that the Coalition’s prowess had fixed the budget emergency. Both arguments support the idea of competence in economic management, but the change in argument was not believable because the earlier argument had become too deeply entrenched in the public consciousness. In promulgating the arguments around a budget emergency so strenuously, the Abbott government painted itself into a rhetorical

12 corner, creating path dependent affects that they could not then tenably extricate themselves from merely by changing arguments.

In measuring the rhetoric of ‘budget emergency’ against the heuristic of variables effecting the path dependent qualities of particular rhetorical choices, all six factors are present, but at varying levels of intensity. Fiscal policy settings of course have high saliency within the electorate. Economic management is frequently cited as one of the key factors determining voting preferences (see Sanders, Clarke, Stewart, and Whiteley 2001). In terms of the proximity to an election, the match is not as acute, nor is the rhetoric presented in the form of an election promise (although the need to cut government spending did underpin many of the promises made). The Abbott-led opposition were characterising Australia as being in a ‘budget emergency’ at least four months prior to the 2013 election, and for almost a year after the election. Whilst the rhetoric had salience, it was not shackled to a particular promise and therefore could not be framed as an election promise that had to be kept.

The length of time for which the rhetoric was pursued diminished the opportunity for it to be presented as contingent on changeable circumstances. Abbott and Hockey stated that the budget was being poorly managed under Labor, and that a better government could and would be able to turn that around. The language chosen was unambiguous, but was not tied to a particular specific outcome. That potentially allowed large amounts of room for rhetorical movement in how an Abbott government might respond to the situation. But it did not provide room for a total reversal of the rhetoric without undermining the authenticity of the original claim. The use of the term ‘emergency’ gave the rhetorical formulation extra emotive force, which helped it gain traction in political discourse as it was repeated and built upon over many months. The media seized on the ‘budget emergency’ terminology with ubiquitous coverage, and it was utilised and repeated extensively both before and after the election. Tellingly, it was then also used as the background for subsequent media stories challenging the sincerity of the government when in early 2015 it moved to change rhetorical path and suggest that the budget emergency was suddenly over.

The case of the Australian budget emergency debate indicates that rhetorical path dependency may have an impact on rhetorical stickiness even where the original formulation is neither specific nor an election promise. Attempting a rhetorical U-turn can have significant political costs associated with it in terms of trust in the ethos of a particular leader, even when the change relates to a quite broad original formulation. Former Prime Minister Kevin Rudd’s rhetorical embrace of climate change as ‘the greatest moral, economic and social challenge of our time’, only to then go on and defer policy action against it, posed similar affects (Grube 2014). Whilst the original rhetoric is broad and leaves room for rhetorical manoeuvre, the strength of the language mitigates against any attempt to subsequently turn away from its core message entirely.

But some aspects of the rhetorical Australian ‘budget emergency’ case did work to mitigate the degree of political cost associated with the dramatic change in message. When the rhetoric was changed, it represented a shift away from a rhetorical stance that

13 had been used to justify deep and heavily contested cuts in government spending. In other words, whilst it was a dramatic rhetorical U-turn, it was towards a formulation suggesting that the budget was now ‘under control’, therefore minimising the need for further spending cuts. It was a rhetoric that allowed for a more populist appeal. The shift was heavily scrutinised in the media, where most commentators questioned whether the new rhetoric was supported by the evidence, so the credibility of the new rhetoric was contested in a way that damaged the ethos of the prime minister and treasurer. But the more populist message embodied within the new rhetoric counteracted that damage, at least in so far as it was reflected in opinion polls.

Rhetorical Path Dependency: Challenges and Opportunities The starting point for this article is that rhetoric matters. It has an enormous impact on contemporary politics, and to study it we need to understand how political arguments actually reflect and shape political ideas, as the work of Finlayson and others suggests. Equally, rhetorical choices have the capacity to shape and re-shape the ideas and traditions that give institutional shape to the political world. In discursive institutionalist terms – drawing on Schmidt – back of mind ideas and front of mind communicative abilities work together to shape understanding through discourse. Individual agency can and does have dramatic impacts in the way that rhetorical choices and formulations shape institutions. Constructing a theory of rhetorical path dependency does not mean that individual agency in making discursive choices is somehow pre-determined by institutionalised boundaries. Rather, it draws attention to the ways that discursive choices made through the free expression of individual agency can subsequently begin to take on institutionalised characteristics that render rhetorical appeals easier to make than to change.

In true path dependent fashion, there are costs associated with exit from one rhetorical path to another, and increasing returns flow to those leaders who are able to be consistent in their rhetoric. The costs are measured in terms of the damage done to the ethos of a leader – the one part of the triptych of Aristotelian rhetorical appeals that can’t be made or re-made through words alone because ethos reflects assessments of the character and trustworthiness of the leader. This damage to ethos may or may not be accompanied by electoral costs in terms of falling opinion polls. Such shifts in electoral support may be heavily influenced as much by the popularity of the new message as by the extent to which it departs from an entrenched rhetorical path.

The tension between rhetorical promise and the difficulties of delivery has existed for decades in democratic politics. Political leaders know that there is a price to be paid for saying one thing and then seeking to justify doing another. Yet political leaders seem to repeat the mistake ad nauseam and it is worth considering why. The reason may be that they actually have little choice. Democratic politics requires leaders to make commitments, to say in a compelling way what they will do should they be entrusted with the government of the people. Engaging in broad hedging rhetoric that allows for complete freedom of action once in government might offer greater flexibility, but it’s unlikely to fulfil the rhetorical need to actually persuade the people to place a particular

14 leader in power. It gives the impression of a leader who either doesn’t know what they stand for, or who stands for nothing and is not willing to make concrete commitments.

The rewards flow instead from promulgating a consistent narrative, with consistent rhetorical commitments to specific policy prescriptions underpinning that narrative. But, faced with an intransigent legislature or an unexpected fiscal squeeze or some other circumstance, a leader who suddenly has to change rhetorical tack will find that their earlier rhetorical formulations stick to them in ways they cannot shake off. Particularly in the modern era of social media and the 24/7 news cycle, every broken promise can be replayed endlessly at the touch of a button, highlighting every rhetorical change of direction in ways that damage the ethos of the political leader by portraying them as inconsistent and untrustworthy. It is part of the nature of political leadership that leaders have to respond to events and changing circumstances, but in changing their rhetoric to do so, there are costs attached to that rhetorical exit.

The cost of that exit is also affected to some extent by questions of timing. Consistently political leaders attempt to brazen their way through the stickiness of earlier promises by not admitting that they have in fact changed, or by simply mounting new arguments for why they had to change. This type of rhetoric simply has the effect of further entrenching the rhetorical path. Realising that they are stuck with their original words, politicians frequently re-commit to them and try and make the policy reality somehow fit with those words. In Australia, the Gillard Government’s commitment to delivering a budget surplus by 2013 is a classic case (see Walter 2014: 43-51; see also Walter and Uhr 2013). Even when it was becoming clear that the government’s taxation income was rapidly diminishing in the face of the mining downturn, the government just kept re-iterating that they would deliver a surplus come what may. They continued to stick to the rhetorical path laid out because the costs of exiting were simply too high to be countenanced in terms of the government’s reputation as a manager of the public finances. This simply operated to deepen the rhetorical commitment and the degree of rhetorical path dependency. When finally faced with the impossibility of delivering a surplus, the government faced extraordinary political pain in having to change rhetorical tack. It mounted new arguments for why it was more responsible to run a deficit in difficult fiscal circumstances, but the new rhetoric remained overshadowed by the old, especially when viewed in the context of decades of debate over whether Labor or the Coalition are the better economic managers (see Fenna 2007).

The study of rhetoric of course has aspects beyond those that are covered here in promulgating a theory of rhetorical path dependency. For example, political leaders must reach out to a range of different audiences in their rhetoric, and this will affect their choice of arguments and how they’re presented. Leaders have different rhetorical roles to play at different times to different audiences (Grube 2013). These are important factors that may influence the overall ‘success’ of a particular rhetorical gambit, even whilst they may have little impact on whether a particular rhetorical formulation takes on path dependent qualities.

Developing and further testing a theory of rhetorical path dependency opens up a range of avenues for future research. Firstly, even where path dependent effects are strongly in evidence, how can political actors seek to counteract those effects? Is it in fact possible to wipe the slate clean and allow a fresh rhetorical start? For example, elections may provide a fresh democratic mandate for leaders to pursue a new direction. Equally, being immediately and openly honest with the electorate when a rhetorical shift is being made may affect how voters will assess the change in direction. Systematic empirical study is needed to assess and analyse such impacts. Another variable that may repay close study is whether the system of government contributes to determining the public’s level of tolerance for rhetorical shifts. For example, in non-majoritarian democracies there may be a greater acceptance that what is said before an election is meant to stake out a bargaining position that will be subject to post-election negotiation with other parties in the quest to form a government.1

Equally important is devising a means of measuring or evaluating the costs of path deviation, and identifying the tipping points at which political actors are willing to pay those costs. Clearly the Abbott government suffered a blow to its ethos for its rhetorical reversal on the language around there being a budget emergency, but that damage did not occur in isolation. It occurred in a context of broader critiques about broken promises and political misjudgements by a government that was consistently performing poorly in the opinion polls. Isolating the extent to which the costs of breaking from a rhetorical path in one policy area contributed to the wider malaise of the government would help to crystallise how far-reaching the damage from rhetorical changes can potentially be. There is also the potential for interdisciplinary conversations between linguistics scholars and political scientists in analysing whether particular linguistic techniques are more likely than others to entrench a political promise in popular discourse.

A theory of rhetorical path dependency opens up these and other research questions by starting with the proposition that political actors can create rhetorical traps for themselves which then take on path dependent qualities. It also highlights the benefits – the increasing returns – that flow to politicians who can be consistent in their rhetorical appeals. And finally, it asks whether – rather than trying to talk their way out of having performed rhetorical shifts – leaders may in fact be better able to protect their ethos by admitting change rather than obfuscating it. Even in rhetorical choices, honesty may in fact be the best policy.

References Abbott, T. 2013. Budget reply speech, Parliament House, Canberra, 16 May 2013. URL: . Consulted 15 July 2015.

1 The author is grateful to Minou de Ruiter for this observation.

Abbott, T. 2015. Transcript of press conference. Parliament House Canberra, 18 March 2015. URL: . Consulted 14 January 2016.

Aristotle, translated by Rhys Roberts, W. 1984. Rhetoric. New York: Modern Library.

Atkins, J. and Finlayson, A. 2013. ‘…A 40-Year-Old Black Man Made the Point to Me’: Everyday Knowledge and the Performance of Leadership in Contemporary British Politics. Political Studies 61: 161–77.

Atkins, J., Finlayson, A., Martin, J. and Turnbull, N., eds. 2014. Rhetoric in British politics and society. Basingstoke: Palgrave Macmillan.

Bell, S. 2011. Do we really need a new ‘constructivist institutionalism’ to explain institutional change? British Journal of Political Science 41: 883-906.

Bennister, M. 2013. Tony Blair’s Oratory. In R. Hayton and A. Crines, eds. Labour party oratory from Bevan to Brown. Manchester: Manchester University Press.

Budge, I. 2015. Issue emphases, saliency theory, and issue ownership: a historical and conceptual analysis. West European Politics 38: 761-77.

Crowe, D. and Uren, D. 2015. Tony Abbott looks at next round of budget cuts, The Australian, 18 March 2015. URL Consulted 17 July 2015.

Eckstein, H. 1975. Case Study and theory in political science. In F. Greenstein and N. Polsby, eds. Strategies of inquiry. London, Amsterdam: Addison Wesley.

Fenna, A. 2007. Governing in good times: fiscal policy and tax reform in Australia 1996- 2006, Australian Journal of Political Science 42: 329-349.

Finlayson, A. 2007. From beliefs to arguments: interpretive methodology and rhetorical political analysis. British Journal of Politics and International Relations 9: 545-63.

Finlayson, A. 2012. Rhetoric and the political theory of ideologies. Political Studies 60: 751-67.

Finlayson, A. and Martin, J. 2008. ‘It ain’t what you say…’: british political studies and the analysis of speech and rhetoric. British Politics 3: 445-64.

Flyvbjerg, B. 2006. Five misunderstandings about case-study research. Qualitative Inquiry 12: 219-45.

Friedman, J. and Friedman, S. 2012. Rethinking the rhetorical presidency. Abingdon: Routledge.

Gains, F., John, P.C., and Stoker, G. 2005. Path dependency and the reform of English local government, Public Administration 83: 25-25.

Grube, D. 2013. Prime ministers and rhetorical governance, Basingstoke: Palgrave Macmillan.

Grube, D. 2014. The gilded cage: Rhetorical path dependency in Australian politics. In J. Uhr and R. Walter, eds. Studies in Australian political rhetoric. Canberra: ANU Press.

Hardin, R. 1982. Collective action. Baltimore: Johns Hopkins University Press.

Hay, C. 2010. Ideas and the construction of interests. In R.H. Cox and D. Beland, eds. Ideas and politics in social science research. Oxford: Oxford University Press.

Hockey, J. 2014. Budget speech 2014-15, Delivered 13 May, Parliament House, Canberra. URL Consulted 14 January 2016.

Hogan, J. 2006. Remoulding the critical junctures approach. Canadian Journal of Political Science 39: 657-679.

Kane, J. and Patapan, H. 2010. The artless art: Leadership and the limits of democratic rhetoric. Australian Journal of Political Science 45: 371-389.

Kay, A. 2003. Path dependency and the CAP. Journal of European Public Policy 10: 405-20.

Kay, A. 2005. A critique of the use of path dependency in policy studies. Public Administration 83: 553-71.

Kernell, S. 2006. Going public: New strategies of presidential leadership (4th edn). Washington DC: CQ Press.

Mackay, F., Monro, S. and Waylen, G. 2009. The feminist potential of sociological institutionalism. Politics and Gender 5: 253-262.

Mahoney, J. 2000. Path dependence in historical sociology. Theory and Society 29: 507- 548.

Mahoney, J. and Thelen. K. 2010. Explaining institutional change: Ambiguity, agency and power. Cambridge: Cambridge University Press.

March, J.G. and Olsen, J.P. 1989. Rediscovering institutions: The organizational basis of politics. New York: The Free Press.

Martin, J. 2015. Situating speech: A rhetorical approach to political strategy. Political Studies 63: 25-42.

Masters, A. and ‘t Hart, P. 2012. Prime Ministerial Rhetoric and Recession Politics: Meaning Making in Economic Crisis Management. Public Administration 90: 759–80.

Medhurst, M. J. ed. 2006. The rhetorical presidency of George H.W. Bush. College Station: Texas A&M University Press.

Naurin, E. 2014. Is a promise a promise? Election pledge fulfilment in comparative perspective, using Sweden as an example. West European Politics 37: 1046-64.

North, D. 1990. Institutions, institutional change and economic performance. Cambridge: Cambridge University Press.

Peters, B. G., Pierre, J. and King, D. 2005. The politics of path dependency: Political conflict in historical institutionalism. The Journal of Politics 67: 1275-300.

Pierson, P. 2004. Politics in time. History, institutions and social analysis. Princeton: Princeton University Press.

Rolfe, M. 2008. New wine into old bottles: Ethical appeals and democratic discourse. Australian Journal of Political Science 43: 513–29.

Sanders, D., Clarke, H., Stewart, M., and Whiteley, P. 2001. The economy and voting. Parliamentary Affairs 54: 789-802.

Schmidt, V.A. 2008. Discursive institutionalism: The explanatory power of ideas and discourse. Annual Review of Political Science 11: 303-26.

Schmidt, V.A. 2010. Taking ideas and discourse seriously: Explaining change through discursive institutionalism as the fourth ‘new institutionalism’, European Political Science Review 2: 1-25.

Schmidt, V.A. 2012. Discursive institutionalism: Scope, dynamics, and philosophical underpinnings. In F. Fischer and J. Forester, eds. The argumentative turn revisited: Public policy as communicative practice. Durham, NC: Duke University Press.

Shepsle, K.A. 2008. Rational choice institutionalism. In S.A. Binder, R.A.W. Rhodes, and B.A. Rockman, eds. The Oxford handbook of political institutions. Oxford: Oxford University Press.

Stuckey, M. E. 2010. Rethinking the rhetorical presidency and presidential rhetoric. Review of Communication 10: 38-52.

Toye, R. 2011. The rhetorical premiership: A new perspective on prime ministerial power since 1945. Parliamentary History 30: 175-92.

Tulis, J. 1987. The rhetorical presidency. Princeton: Princeton University Press.

Uhr, J. and Walter, R. eds. 2014a. Studies in Australian political rhetoric. Canberra: ANU Press.

Uhr, J. and Walter, R. 2014b. Studying Australian political rhetoric. In J. Uhr and R. Walter, eds. Studies in Australian political rhetoric. Canberra: ANU Press.

Uhr, J. and Walter, R. 2015. The rhetorical standards of public reason in Australia. Australian Journal of Politics and History 61: 248-262.

Walter, R. 2014. Economic management, rhetorical tactics, and the cost of promises. In J. Uhr and R. Walter, eds. Studies in Australian political rhetoric. Canberra: ANU Press.

Walter, R. and Uhr, J. 2013. Budget talk: Rhetorical constraints and contests, Australian Journal of Political Science 48: 431-444.

Young, S. 2007. Political and parliamentary speech in Australia. Parliamentary Affairs 60: 234-52.

Figure 1: Heuristic of Variables that Shape Rhetorical Path Dependent Effects.

Contributing Factors Impact Proximity to an election Rhetoric close to an election is likely to be considered an ‘election promise’, increasing its path dependent effects.

Political saliency Rhetoric in areas that are central to the concerns of voters is likely to have higher path dependent effects than rhetoric on issues of low political saliency. Level of contingency and control The less contingent on circumstances a promise is, the higher the path dependent effects are likely to be because political actors will be held to be in control. Specificity of the language in the original Unambiguous language is likely to lead to rhetoric higher path dependent effects. The more emphatic a piece of rhetoric, and the ‘Depth’ of the rhetoric more often it is repeated over time, the higher the path dependent effects are likely to be. Media Coverage The more media coverage a piece of rhetoric receives, and the number of news cycles for which it does so, the higher the path dependent effects are likely to be.